we should only keep around uvdata objects with valid vis_units
aewallwi opened this issue · 0 comments
aewallwi commented
mK are not a valid visibility unit for a UVData object (valid units are either Jy or K sr, see https://github.com/RadioAstronomySoftwareGroup/pyuvdata/blob/9ab7332cabd065aac2ebf0931c0ed7dc02cabc89/pyuvdata/uvdata/uvdata.py#L84), but in hera_pspec, we like to carry around uvdata objects with units of mK which are technically not valid units for this data type.
We should organized hera_pspec normalization in a way that does not require carrying around invalid uvdata objects.